Gemma 3 12B vs Hy-MT2-1.8B

Prices checked

Gemma 3 12B takes 4 of 5 rounds against Hy-MT2-1.8B. It is also the cheaper of the two to run, at $0.06 per million tokens against $0.08.

GoogleWins 4 rounds

Gemma 3 12B

131k context · tool calling · reads images · open weights

$0.06/M tokens

Cheapest price, at Kilo Gateway · 3.1¢ per 1,000 pages

Tencent

Hy-MT2-1.8B

8k context · open weights

$0.08/M tokens

Tencent's own price · 3.7¢ per 1,000 pages

Maker's price in (lower wins)

$0.05/M

$0.04/M

Maker's price out (lower wins)

$0.15/M

$0.18/M

Cheapest price (lower wins)

$0.06/M at Kilo Gateway

none below the maker

Maker's price, blended (lower wins)

$0.08/M

$0.08/M

Cost per 1,000 pages (lower wins)

3.1¢

3.7¢

Context

131k tokens

8k tokens

Open weights

yes

yes

Tool calling

yes

no

Structured output

yes

no

Reads images

yes

no

Reasoning

no

no

Released

13 Mar 2025

20 Aug 2026

Providers selling it

16

4

Price we measure against

OpenRouter

the maker's own

Trains on your prompts (no is better)

not published

no provider to check

Providers selling both · 3

  • Kilo Gateway$0.06 / $0.08
  • OpenRouter$0.07 / $0.08
  • AI/ML API$0.08 / $0.08

Gemma 3 12B, then Hy-MT2-1.8B, per million tokens.

Go further

Add a third or fourth model, or change the weighting, in the tool.

Frequently asked

Which is cheaper, Gemma 3 12B or Hy-MT2-1.8B?
Gemma 3 12B, at $0.06 per million tokens against $0.08 for Hy-MT2-1.8B, blended 3 in : 1 out, read on 31 Oct 2026.
Where is each one cheapest?
Gemma 3 12B: $0.06 at Kilo Gateway. Hy-MT2-1.8B: nobody sells it below the maker's own price with a stated reason, so $0.08 is the number to beat.
Can I buy both from one provider?
Yes. 3 providers sell both: Kilo Gateway, OpenRouter, AI/ML API. The cheapest of them for the two together is Kilo Gateway.
Which holds more context?
Gemma 3 12B, at 131k tokens against 8k.

Prices exclude tax, read 31 Oct 2026 from each provider's own catalog. They are per M tokens, blended 3 in : 1 out, and a page is 600 tokens in and 60 out. Bars are scaled to the larger value in each row.

Related comparisons

More